From time to time, computing devices may require system-level changes to maintain proper and/or efficient operation. Examples of such changes may include changes to the BIOS, patches for the processor microcode, firmware updates, security patches, configuration changes, and/or others. However, implementing a system change may cause the computing device to enter an inoperable state. As a result, performing a system-level change in an automated (e.g., unattended) fashion can be risky. In fact, if a computing device receives a request for an automated system change, the request may be ignored or denied so as not to risk a system failure, despite the potential benefits of installing the system change.